Mechanism to adapt garbage collection resource allocation in a solid state drive
Methods and apparatus related to a mechanism for quickly adapting garbage collection resource allocation for an incoming I/O (Input/Output) workload are described. In one embodiment, non-volatile memory stores data corresponding to a first workload and a second workload. Allocation of one or more resources in the non-volatile memory is determined based at least in part on a determination of an average validity of one or more blocks, where the one or more candidate bands are to be processed during operation of the first workload or the second workload. Other embodiments are also disclosed and claimed.
1. An apparatus comprising:
non-volatile memory to store data corresponding to a first workload and a second workload; and
logic to determine allocation of one or more resources in the non-volatile memory based at least in part on a determination of an average validity of one or more blocks to be processed during a transition from the first workload to the second workload.
2. The apparatus of claim 1 , wherein the logic is to determine the allocation of the one or more resources to garbage collection logic and a host coupled to the non-volatile memory.
3. The apparatus of claim 2 , wherein the garbage collection logic is to free space occupied by invalid data in the non-volatile memory.
4. The apparatus of claim 2 , wherein the logic to determine the allocation of the one or more resources is to comprise the garbage collection logic.
5. The apparatus of claim 1 , wherein logic to determine the allocation of the one or more resources in the non-volatile memory based at least in part on the determination of the average validity of the one or more blocks to be processed during operation of the first workload or the second workload.
6. The apparatus of claim 1 , wherein the logic is to determine the allocation of the one or more resources to cause an increase in an effective spare space of the non-volatile memory.
7. The apparatus of claim 1 , wherein the logic is to determine the allocation of the one or more resources to cause a reduction of write amplification in the non-volatile memory.
8. The apparatus of claim 1 , wherein the second workload is to immediately follow the first workload.
9. The apparatus of claim 1 , wherein the first workload is an empty or idle workload.
10. The apparatus of claim 1 , wherein the non-volatile memory and the logic are on a same integrated circuit device.
11. The apparatus of claim 1 , wherein the non-volatile memory is to comprise one of: nanowire memory, Ferro-electric transistor random access memory (FeTRAM), magnetoresistive random access memory (MRAM), flash memory, Spin Torque Transfer Random Access Memory (STTRAM), Resistive Random Access Memory, Phase Change Memory (PCM), and byte addressable 3-Dimensional Cross Point Memory.
12. The apparatus of claim 1 , wherein an SSD is to comprise the non-volatile memory and the logic.
13. A method comprising:
storing data corresponding to a first workload and a second workload in a non-volatile memory; and
determining allocation of one or more resources in the non-volatile memory based at least in part on a determination of an average validity of one or more blocks processed during a transition from the first workload to the second workload.
14. The method of claim 13 , further comprising determining the allocation of the one or more resources to garbage collection logic and a host coupled to the non-volatile memory.
15. The method of claim 13 , further comprising the garbage collection logic freeing space occupied by invalid data in the non-volatile memory.
16. The method of claim 13 , wherein determining the allocation of the one or more resources in the non-volatile memory causes an increase in an effective spare space of the non-volatile memory.
17. The method of claim 13 , wherein determining the allocation of the one or more resources in the non-volatile memory causes a reduction of write amplification in the non-volatile memory.
18. The method of claim 13 , wherein the first workload is an empty or idle workload.
19. The method of claim 13 , wherein the non-volatile memory comprises one of: nanowire memory, Ferro-electric transistor random access memory (FeTRAM), magnetoresistive random access memory (MRAM), flash memory, Spin Torque Transfer Random Access Memory (STTRAM), Resistive Random Access Memory, Phase Change Memory (PCM), and byte addressable 3-Dimensional Cross Point Memory.
20. The method of claim 13 , further comprising determining the allocation of the one or more resources in the non-volatile memory based at least in part on the determination of the average validity of the one or more blocks processed during operation of the first workload or the second workload.
21. A system comprising:
non-volatile memory; and
at least one processor core to access the non-volatile memory;
the non-volatile memory to store data corresponding to a first workload and a second workload; and
logic to determine allocation of one or more resources in the non-volatile memory based at least in part on a determination of an average validity of one or more blocks to be processed during a transition from the first workload to the second workload.
22. The system of claim 21 , wherein the logic is to determine the allocation of the one or more resources to garbage collection logic and a host coupled to the non-volatile memory.
23. The system of claim 21 , wherein the logic is to determine the allocation of the one or more resources to cause an increase in an effective spare space of the non-volatile memory.
24. The system of claim 21 , wherein the logic is to determine the allocation of the one or more resources to cause a reduction of write amplification in the non-volatile memory.
25. The system of claim 21 , wherein the first workload is an empty or idle workload.
26. The system of claim 21 , wherein logic to determine the allocation of the one or more resources in the non-volatile memory based at least in part on the determination of the average validity of the one or more blocks to be processed during operation of the first workload or the second workload.